关键词: 多核计算，图像稳定，运动估计，并行处理

Abstract: The multi-core computing has become the new trend of computer industry. Electronic Digital Image Stabilization (EDIS) system involves both large amount of data and intensive computation. The real-time implementation of EDIS system entails rapid data processe with high parallelism and repeated computation,which are suitable for data parallel processe of image sectuences. The algorithm solution was proposed to deal with high speed image stabilization to meet the requirements of real-time systems. According to the data processe characteristics of EDIS, efficient C++ with inline SIMI)(Single Instruction Multiple Data) and multi-thread programming were written and compiled on dual-core PC for real parallel running threads to achieve real-time performance in conventional PC. After the motion vectors acquired from fast search strategy, adaptive multi-local motion vectors avoided the moving objects interference. And the Kalman filter removed high-frequency image fitter motion while smooth global movements were retained for scanning and vehicle moving. The test results show that it is possible to implement a high efficient and robust real-time stabilization system on conventional PC with Multi-core at last.

Key words: Multi-core computation, Image stabilization, Motion estimation, Parallel processing